Genetic And Environmental Risk Factors For The Development Of Pulmonary Tuberculosis In Patients With Type 2 Diabetes Mellitus

Objective: To explore environmental risk factors for the development of pulmonary tuberculosis in patients with type 2 diabetes mellitus(T2DM-PTB). Method: A hospital-based case-control study was conducted at two hospitals in Tianjin during March 2005 to March 2006. 158 newly-diagnosed PTB patients who had had type 2 DM were recruited as cases, and 281 type 2 DM patients were choosen as the control group which was matched by age and sex in frequency to the case group. Information on risk factors was collected through face-to-face interview. Unconditional logistic regression model was used to analyse the risk factors for T2DM-PTB. New confounding caused by matching on age and sex was excluded by stratified analysis. Multivariate logistic regression modified by the principal component analysis, factor analysis and K-Mean Cluster analysis were conducted and final main effect functions were built by different ages and sexes. Three kinds of method were used to explore the potential interactions among risk factors: interaction analysis, adding interaction items to the multi-variate logistic regression model and general relative risk model.Results: 21 variables were found to be associated with T2DM-PTB in the univariate analysis, which were education, income , occupation, blood sugar control, diet control after DM diagnosis, weight loss, maximal BMI, stroke, hypertension, TG, labor intensity, physical exercise, smoking, alcohol drinking, tea drinking, exposure to dust,room size,house lighting,house ventilation,personal character and life event. The modified multivariate logistic regression analysis showed that low socioeconomic level, bad housing condition, bad compliance toDM therapy, smoking and drinking were significantly associated with increased risk of T2DM-PTB in the main effect model, and the OR were 3.308(95%CI: 2.414— 4.535), 2.830(95%CI: 2.138—3.747), 1.814(95%CI: 1.407-2.339) and 1.647 (95%CI: 1.278—2.121) respectively;while tea drinking and HDL. hypertention and obesity, stroke were shown to be reversely associated with T2DM-PTB and therefore might prevent from the development of T2DM-PTB and the OR were 0.643(95%CI: 0.498-0.830), 0.640(95%CI: 0.493—0.831) and 0.552(95%CI: 0.417—0.730) respectively.Furthermore, risk factors for different sex and age groups were explored and found that:Men: The risk of T2DM-PTB among men increased with low socioeconomic level, bad housing condition, bad working condition n bad compliance to DM therapy and the OR were 4.497(95%CI:2.838—7.127). 3.028(95%CI:2.083—4.401) 1.690(95%CI: 1.245—2.295) and 1.470(95%CI: 1.076-2.008) respectively;while stroke(OR=0.675,95%CI: 0.456—0.999), extroversion(OR=0.553, 95%CI:0.386~ 0.792). hypertention and obesity(OR=0.510, 95%CI:0.348—0.748) might reduced the risk of T2DM-PTB among male population.Women: The OR for T2DM-PTB raised among those women who with low socioeconomic level(OR=4.860, 95%CI:2.468 — 9.569) > bad housing condition (OR=3.882,95%CI:2.169—6.947). smoked and drinked (OR=2.883,95%CI: 1.002-8.294) and bad compliance to DM therapy(OR=2.678, 95%CI:1.588—4.517);while those with hypertention and obesity(OR=0.629, 95%CI:0.415—0.953). tea drinking habit and higher HDL(OR=0.549,95%CI:0.351—0.859). stroke(OR=0.394, 95%CI: 0.246—0.632) had the lower risk of T2DM-PTB in women.Age less than 60: In younger diabetics, those who with low socioeconomic level(OR=5.568, 95%CI: 3.113—9.957). bad housing condition (OR=2.987, 95%CI: 1.967—4.535). smoked and drinked (OR=1.539, 95%CI: 1.026—2.309). badcompliance to DM therapy(OR=1.913, 95%CI: 1.317—2.781) might have a higher risk of T2DM-PTB;while tea drinking and higher HDL(OR=0.575, 95%CI:0.405— 0.817), stroke(OR=0.564, 95%CI:0.340 - 0.937) might reduce the risk of T2DM-PTB in the relatively yonger population .Age more than 60: For old diabetics, bad housing condition (OR=3.325, 95%CI: 2.150-5.141), low socioeconomic level(OR=3.207, 95%CI: 2.057-4.998), bad compliance to DM therapy(OR=2.282, 95%CI: 1.522—3.420) , smoked and drinked (OR=1.733, 95%CI: 1.229—2.443) increased the risk of T2DM-PTB;while hypertention and obesity(OR=0.501, 95%CI:0.328—0.766), stroke(OR=0.481, 95%CI:0.328—0.705) , DM status(OR=0.678, 95%CI:0.456— 1.007) might decrease the risk of T2DM-PTB among old people.The results of interaction analysis showed that no additive-model intercation was found between the risk factors. The significant interaction introducing to the multi-variate logistic regression model was multiplication of bad control of blood sugar and diet. GRRM analysis indicated that the pattern of interaction between the risk factors of the data tended to be multiplicative model and using Logsitic regression to analysis of the interaction was correct.Conclusion: There was a significant association between T2DM-PTB and low socioeconomic level;bad housing condition or bad compliance to the treatment of DM in the all sex and age groups;smoking and drinking might increase the risk of T2DM-PTB in in all statified groups except in stratum of male. While in male diabetics, bad working condition might have a higher risk for T2DM-PTB. The distribution of protective factors in different strata was varying.The risk factors or protective factors were different according to different sex and age, and these factors might cooperate with each other to influence the occurance of T2DM-PTB.